Skip to Main content Skip to Navigation

Heliosat-V ˸ une méthode polyvalente d’estimation du rayonnement solaire au sol par satellite

Abstract : Solar irradiance at the surface of the Earth is recognized as an essential climate variable by the World Meteorological Organization. Its knowledge is as much important for climate sciences as for the development of energy alternatives to fossil fuels, like solar photovoltaic. Ground measurements of this radiation are very sparse on Earth, explaining the interest for satellite-based remote sensing to estimate it. Combining estimations from different satellites in orbit is a pathway to cover the information on the whole globe. Different generations of satellites also produced a multidecadal imagery of the Earth, making it conceivable to estimate long time series of solar radiation, or even to identify long-term variations, a recurrent objective in the study of climate change. For more than 30 years, Heliosat methods estimate surface solar irradiance from satellite imagery, but they have been designed to be applied to a specific sensor on a geostationary orbit, and have limitations in their scope : the need for a long archive of satellite imagery (Heliosat, Heliosat-2), or else the need for multispectral measurements (Heliosat-4). This work dedicated to the development of a Heliosat-V method brings elements of versatility to the satellite-based estimation from so-called "cloud-index" methods, with the ultimate goal to reach homogeneous data of solar radiation derived from measurements made by different satellite instruments. Two issues are in particular considered here to reach such an estimation: the diversity of sensors in terms of spectral sensitivities, and the influence of viewing and solar geometries on spaceborne measurements. The method extensively deals with radiative transfer modeling in the spectral range 400-1000 nm to simulate on one hand satellite measurements in clear-sky conditions, and on the other hand satellite measurements in the presence of an optically thick cloud. The method is tested on the imagery of a geostationary satellite instrument, Meteosat-9/SEVIRI, and in a more exploratory way, on the non geostationary sensor DSCOVR/EPIC. Results are compared to high quality ground-based measurements of irradiance, and show performances similar to operational satellite products. However, the quality of estimates depends on the spectral channel used, and especially of the presence of clear-sky atmospheric scattering or absorption in the signal measured by the satellite instrument. The accent is also put on the need for an accurate absolute calibration of satellite radiometric measurements in order to produce time series of surface solar irradiance with the smallest biases and temporal drift possible.
Complete list of metadata
Contributor : ABES STAR :  Contact
Submitted on : Monday, May 17, 2021 - 10:33:27 AM
Last modification on : Thursday, April 28, 2022 - 11:28:51 AM
Long-term archiving on: : Wednesday, August 18, 2021 - 6:16:31 PM


Version validated by the jury (STAR)


  • HAL Id : tel-03227271, version 1


Benoît Tournadre. Heliosat-V ˸ une méthode polyvalente d’estimation du rayonnement solaire au sol par satellite. Génie des procédés. Université Paris sciences et lettres, 2020. Français. ⟨NNT : 2020UPSLM063⟩. ⟨tel-03227271⟩



Record views


Files downloads